tensorchord/modelz-llm

Issues

• Local custom .gguf modals supported?

  #102 opened 8 months ago by sengiv
  2

• Tokenizer class LLaMATokenizer does not exist or is not currently imported.

  #101 opened 9 months ago by ken-xu-e
  1

• Cant use it on Windows

  #98 opened 9 months ago by yogeshhk
  3

• Missing a LICENSE

  #99 opened 9 months ago by loleg
  1

• bug: Failed to generate outputs

  #97 opened a year ago by gaocegege
  2

• Function calling feature

  #96 opened a year ago by willswordh
  1

• add llama-2

  #90 opened a year ago by antonkulaga
  1

• do we support vicuna 13b, chatglm2 ?

  #88 opened a year ago by timiil
  1

• feat: provide instructions on how community members can wrap models for this project

  #87 opened a year ago by PaulConyngham
  1

• bug: can only concatenate str (not "int") to str

  #86 opened a year ago by gaocegege
  1

• bug: Completion request returns wrong response

  #83 opened a year ago by gaocegege
  0

• chore: Fix vicuna 7b

  #82 opened a year ago by gaocegege
  0

• feat: support chatgpt web

  #79 opened a year ago by dgqyushen
  1

• feat: Support falcon 7b

  #77 opened a year ago by gaocegege
  0

• bug: Unexpected OOM in ChatGLM 6B

  #69 opened a year ago by gaocegege
  2

• feat: Support more models

  #11 opened a year ago by gaocegege
  3

• bug: Vicuna performance is not great

  #73 opened a year ago by gaocegege
  2

• bug: RuntimeError: "LayerNormKernelImpl" not implemented for 'Half' in chatglm int4

  #62 opened a year ago by gaocegege
  2

• bug: AttributeError: 'ChatGLMForConditionalGeneration' object has no attribute 'encoder'

  #63 opened a year ago by arugal
  4

• bug: Install cuda again in the image

  #55 opened a year ago by gaocegege
  2

• bug: RuntimeError: Only Tensors of floating point and complex dtype can require gradients

  #60 opened a year ago by gaocegege
  0

• feat: Match lmsys/fastchat-t5-3b-v1.0 performance with fastchat

  #42 opened a year ago by gaocegege
  2

• bug: Object missing required field `model`

  #50 opened a year ago by gaocegege
  0

• test: check if the embedding API is compatible

  #45 opened a year ago by kemingy
  0

• bug: Extra blank space in output

  #48 opened a year ago by gaocegege
  0

• feat: Support flag to use CPU/GPU

  #13 opened a year ago by gaocegege
  0

• feat: Support CICD for different model dry-run warmup

  #9 opened a year ago by gaocegege
  0

• feat: Add a more elegant way to specify the model

  #10 opened a year ago by gaocegege
  2

• feat: Add CLI argument int8 and int4

  #38 opened a year ago by gaocegege
  0

• bug: 500 when exceeds the max token length with chatglm 6b int4

  #35 opened a year ago by gaocegege
  0

• feat: Support embedding API

  #14 opened a year ago by gaocegege
  2

• bug: 500 with langchain sdk

  #16 opened a year ago by gaocegege
  2

• feat: Remove v1 from URI

  #8 opened a year ago by gaocegege
  0